HP-65 Program "persists" and steps "disappear"

[n1msr]

Just had an HP-65 revived with one of the teenix CPU boards plus a card reader and LED repair.

The calculator seems to work fine if I read in a program from a magnetic card, or transfer a program via Bluetooth that the teenix CPU board offers (the original CPU board failed).

But manual entry seems to be an issue. The symptom observed is described below, but it might be the W/PRGM <--> RUN switch that is the cause.

Today I was typing in programs from the Standard Pac because some of the cards I have for the HP-65 it can no longer be read. This is when I came across some strange behaviour:

1. I type in the first 4 steps of program STD 01A (personal investment):

STO 1
STO 2
STO 3
R/S

I then go to RUN mode, hit RTN to go back to the start of program memory. Press a numeric key then R/S to run the program. I check the R1, R2 and R3 registers. Only R3 has the number stored. (R1 and R2 are zero). Next...
I hit RTN to go to start of program memory
Switch to W/PRGM mode and SST
The first instruction is STO 3. The first two steps (STO 1 and STO 2) have gone.

Also, when I clear the Program, turn off the calculator and turn it on again, there is a different (previous) set of program steps in the program memory starting LBL A. Weird.

The only other symptom I have observed is when switching between W/PRGM and RUN mode, there is a flash in the LED display. I am not sure if this is affecting the memory.

I have contacted teenix.org and will be checking the switch contacts. But wanted to post this here in case another has seen this kind of thing?

Regards,

Mark.

[Dave Britten]

(01-26-2024 03:17 PM)n1msr Wrote:  Also, when I clear the Program, turn off the calculator and turn it on again, there is a different (previous) set of program steps in the program memory starting LBL A. Weird.

This sounds like normal behavior. When powering on the 65, the A-E keys are given their default functions (1/x, square root, etc.) by way of some small default programs loaded into memory. When entering a new program, always clear program memory first, even when you've just powered on.

I don't know if that has anything to do with the behavior you're seeing with regards to entering the personal investment program. If I power on my 65, switch immediately to W/PRGM, and key in STO 1, STO 2, STO 3, R/S, it appears to retain all four steps, with the default programs (beginning with LBL A) simply pushed down a bit in memory. Of course, this is on an original 65 - I don't know if the Teenix board has any behavioral differences, by design or otherwise.

[jftman]

I think a step RTN usually goes at the end of a program.
RTN inplace of R/S
But if it isn’t there could your test be the bug?

[n1msr]

Thank you both. I am learning!

The 65 is different from the 67 in several ways. I need to download the User Guide and read.

Mark

[Dave Britten]

(01-26-2024 10:59 PM)n1msr Wrote:  Thank you both. I am learning!

The 65 is different from the 67 in several ways. I need to download the User Guide and read.

Mark

Yeah, there are some surprising differences from models that came after (such as tests skipping two steps when false, because GTO is unmerged). Also, I believe RTN has no effect if you haven't started a program/subroutine via A-E. Many of these peculiarities can be exploited to your advantage when writing programs. I would recommend reading through some of the early issues of HP 65 Key Note and PPC Journal if you have access to them, as many interesting tips and tricks are discussed.

EDIT: I forgot to mention that Gene Wright's old web site has some nice games for the 65, as well as an excellent TVM program (actually two programs - one to solve for i% and one to solve for any of the other four variables). These can be good practice for keying in programs, and are worth saving to mag cards.

http://rskey.org/gene/hpgene/index.html

[jftman]

Yeah like written magnetic cards made on a HP 65 can not be read into a HP 67/97.

And the other way too I suppose.
Kind of creates a market for program pacs!

[John Garza (3665)]

(01-27-2024 07:58 PM)jftman Wrote:  Yeah like written magnetic cards made on a HP 65 can not be read into a HP 67/97.

And the other way too I suppose.
Kind of creates a market for program pacs!

Just because they use the same physical media does not imply compatibility. The data is recorded in different formats and density. It wasn't until the 41C that we had developed the ability to read in a previous calculators mag cards.

-J

[John Garza (3665)]

And one of my favorite oddities on the 65...

You don't have to use LBL A to have a label A in your program. Saves a couple of steps!

-J

[John Garza (3665)]

A pearl of wisdom...

Most revolutionary works of engineering (the HP-65 surely qualifies) are idiosyncratic and have faults/oddities. It's usually left to the next version to clean up the mess. The 67 is a much cleaner and well-ordered machine than the 65. Same can be said of the 35/45, the 41/42, etc.

-J

[teenix]

Some new parts arrived today and I was able to build a 65 board. Yes, the mentioned fault did exist. There was a PIC instruction missing, not quite sure why, but without it, the memory delete function got called twice instead of once. I replaced the missing code and now it works ok. New upload available at teenix.org

cheers

Tony